Is this live yet Gus?
In my view, it is a way for quickly and inexpensively developing a series of prototypes, to validate a product idea from a technical and/or market perspective (“will it work that way?” and “will customers buy something like this?”). Creating more than one prototype is usually not an option for embedded projects, but with something like Gadgeteer, it becomes realistic.
If the end product will be produced only in small numbers (let’s say a couple of hundred units or fewer) [em]and[/em] they are uncritical in terms of mechanical properties, electrical properties, certifications etc., then it can make sense to use the same hardware for the products as for prototyping. Otherwise, a custom PCB is the way to go.
In any case, I’d minimize the number of layers, so although I like the Gadgeteer hardware standard (at least the simpler socket types), I never use the Gadgeteer software layer for production code.
There are further ways for optimizing a design, e.g. rewriting part of the code in C, if and where this makes sense. In particular, if it allows lower unit costs (use a slower and thus less expensive microcontroller). But here we are talking about tens of thousands of units to make it worthwhile.
While prototyping is an important activity, I often see that the remaining 90% or so of product development (and longer term product management over several versions) is grossly underestimated. An obsolescence management strategy is certainly part of all that. The nice thing about NETMF is that it makes moving from one hardware to another far easier than with most embedded software platforms. So NETMF can actually be part of an obsolescence strategy. Although the devil’s in the details here as well: porting NETMF to a completely new processor architecture is a huge undertaking, while porting to another board with the same microcontroller is usually easy. Or with the same SOM.
As Gus has said, longer-term availability of a SOM (or single-chip microcontroller) is the really critical thing. I’d expect to tweak the custom carrier board every year or so.
4 Likes
Here I assumed that the SOM [em]will[/em] remain available. But some parts on the carrier board may not be available anymore, or there is some minor change in the requirements, etc.
It’s just part of a product’s planning and budgeting process in a world full of uncertainties. If a year goes by and no changes were necessary, all the better! Bring out the beer 
@ Dave McLaughlin - There will be an announcement when it’s ready.
@ Josh/Gus - So, ummm, can someone at GHI add the Panda 2 so that sexy beast can be revived?
Yea, I want to add the L6470 board with some updates.
Can we have the Domino too ? 
Perhaps I have a few USBizi144’s in a tray somewhere that I could deploy… or perhaps I should do the “breakout board” equivalent I’ve threatened to do for way too long and never got around to it…
Is there a hidden message behind those (wise) requests ? :think: